This isn’t anything new and I feel no ownership over the ideas. I did though, come to my own understanding of these principles and it changed how I perceive game making.

It started because of my random interest in random stuff… I was reading about story development (as in writing stories) and I came across Ernest Hemingway’s theory called the Iceberg Theory. As I understand it, the writer in order to make a good story, should do the following thing: Only expose to the reader the amount of information he needs in order for him to enjoy and make sense of the story. This means it is not necessary to show and explain everything. For example the story “Old Man and the Sea” on the surface level (what is actually written) is about a fisherman struggling with a big fish. But what is actually going on in the readers mind is an epic struggle of the man himself, his age, weaknesses and the heartless forces of life & death.

Next thing I learned about was Chekhov’s gun. This is a literary technique popularized by Russian short-story author Anton Chekhov. The technique states that no element in a story should be introduced if it is not used later in the story. Or in other words: everything that is in the story must have some use later on. The gun was actually mentioned early in the novel “Uncle Vanya”. It is mentioned as an insignificant prop. Later on however it is used as a key plot element as a homicide weapon.

I could call merging these two techniques a Zen approach to writing. Meaning, there is nothing too much in the story. Everything fits in the story and has its use. And on the other hand there is nothing too less. There is always enough written so that your mind will make sense of the story and its flow.

Paraphrasing with how Albert Einstein would say it:

The next thing that lead me to my principles is an article that I have posted on this blog. It’s called Less Talk More Rock. I don’t think you should take this too literally (well, nothing should) but the bottom-line of the article is this. In any case you don’t have to use words or text, don’t use them and actually use the thing. Saying it another way: don’t write ‘rock’. Throw the rock and make it hurt!

Again this is a Zen concept. If you would ask a Zen Master “what is reality?” he wouldn’t reply with any word. He might for example hit his gong: BOOIINNG!! Because nothing you say will represent reality, those are just words! Symbols in your head. Reality is the actual sound, physically going out of the gong and hitting your eardrums, and the experience of it.

So the idea behind Less Talk More Rock is not to describe an experience to a player by (for example) writing text that introduces the player into the game world. Give him the actual experience instead. Hit the gong so that he hears it!

The most brilliant use of this principle is how the game “Another World” starts. If you haven’t played it, go play it. Or watch a video. The intro does not have any text. Anyone, no matter what language he speaks, knows what is going on. It is a scientist, working on some top secret stuff getting blown out into a different planet. The game also seamlessly blends the intro with the actual start of player control (you must swim to the surface from underwater).

Now this cannot be achieved without the use of 3 tools. These will be inventions, at least my understanding of general game design principles. The first tool I called Atoms. The second are Archetypes. Third is Anton’s gun.

The Atom is a an atomic gameplay element. In “Another World” the Atom in the beginning is simply the possibility to swim left, right, up, down inside the water. When you get out you can run left or right and jump up. The movement here is atomic. Meaning you cannot divide this into simpler elements without losing something. Remember: “Everything should be made as simple as possible, but no simpler”. So we could for example take out a dimension from the games movement and make the character only walk left or right. But that wouldn’t actually feel enjoyable because we ourselves do not move like that. That is too simple. We move in 3 dimensions. So left/right + jumping just feels right.

So an Atom is an indivisible gameplay mechanic that feels just right. Most 80’s classics are solely based on this principle. Actually those games invented it. Every other game in the future (every successful game) just copies those elements, mixes them differently and wraps them in different packaging.

Pong

Can Pong be any simpler? Can it be more complicated? Yes it can be, but will that make the game any better, more fun? Never. Expanding the Atom never works. Meaning, it is never significantly better. Only adding another Atom and joining it with another would make the game better. So I will deliberate that adding a second ball to Pong would not make the game better. Because that is the same mechanic of bouncing the ball just more complicated. But introducing a new Atom like for example a block that can be destroyed will make it a totally different, much more fun experience. This way we just invented the game Breakout.

How do you make Breakout better? For example adding another dimension and calling it Breakout 3D? That is an actual game and it is not significantly better than Breakout in my opinion. 3D is just complicating stuff not making it more fun. Instead of expanding the gameplay element you add a new atomic one. For example you add blocks that go at you and hurt you. These are enemies. The rest is just the map boundaries. We now have 3 Atoms:
1. Movement
2. Shooting balls
3. Destroying/Escaping enemies

What game might this be? For example River Raid.

Other Atoms we can build with are: Hitting stuff; Chasing/Running away; Physics rules; Dialogs; Crafting objects; Unlock puzzles (as in find key to open door) and so on.

This is as good as it gets. You can’t reinvent the wheel. Having 2-4 elements like this is sufficient to have a hit game. If you start adding more you risk not having a clear Atom and players will be confused and not enjoy the game. 80’s games and today’s casual games rely solely on 1 Atom. Most super commercial games (like Farmville) rely on 1 to reach the largest audience.

Exercise: What are the Atoms in Soldat?

The next thing I want to explain are Archetypes. This has to do with the wiring of our brains (something Carl Jung found out about). I have a different game design twist on this.

We humans enjoy immensely models. We love to play with models. Actually playing and interacting with models is the same thing. What I mean by models is stuff that represents reality. Every toy you had as a child was a model of something in the real adult world. Lego blocks are basic construction elements. They represent real building construction blocks. Yet children do not play with the real thing. It simply wouldn’t be fun. Our brains are wired to derive pleasure from interacting with simple models.

When I thought about this I arrived at the following conclusion. The more something is realistic, the less it is like a model. Hence it is less fun. Why did I have immense satisfaction playing the game Airborne Ranger where the player was about 11 pixels and all you could do is move up and shoot? Whereas I play a Modern War-type of game where every detail is simulated and my only reaction is blehh, yawn.

I think the actual enjoyment is created in the mind. Simply your mind adds elements that are not there and it creates the game in your imagination. And your imagination is far better than any game graphics and physics realism could ever be. When I look at the 11 pixel soldier my brain knows it looks like shit, so instead of thinking about it in real terms it thinks about it in terms of a model it is playing with.

Realistic looking games start a process in my head where my brain is fooled that it is reality. So I begin interacting with the game as if it was real:
– I test the boundaries.
– I try to kill a civilian.
– A text appears I cannot shoot civilians.
– I get mad because I got fooled thinking this is reality.
– I proceed to Control Panel to uninstall the game.

How do you create an Archetype? It is not a matter of limiting the amount pixels. Remember: “Everything should be made as simple as possible, but no simpler”. There is no need for a lot of pixels and high polycount but at the same time there must be enough pixels so that you know it is a soldier.

Also the Archetype must be something that your brain can associate with. Either in the real world around you or from your childhood memories. A soldier is an archetype because you’ve watched many movies about them. When you see a couple pixels with a gun you know what it is. You know what to expect from it. It should run and shoot. If the soldier does not run and shoot and instead has conversations when FIRE is pressed you get confused and you delete the game.

An Archetype must behave as you expect the thing it represents to behave. An Archer shoots arrows. A Knight strikes his sword. A Mage casts a spell. A Goblin attacks you. A Car drives. A Machine Gun shoots. A bigger Machine Gun shoots more powerfully. A Platform is jumpable. A Ladder lets you walk up and down. Again most 80’s games are made entirely out of Archetypes.

80’s games are also built using the Chekhov’s gun technique. In Anton Chekhov’s words:

Think of Manic Miner.

No element in there is not of some use. There are no decorations or things to set you in the “mood”. It has platforms, ladders and stuff that kills you. There is no element there to confuse you. You cannot break a boundary of a game like this because everything in it works as it should. In a modern game there might be a ladder but it is just a prop made by the map artist. You cannot walk on it. Why is it there? So that it can look cool on a screenshot. But the actual gameplay experience is just frustration when you see there is an invisible collision box around it.

Everything in a game should resemble and behave like the thing it represents. Those are my Archetypes.

If the element in the game has no use to the player it is redundant. That is Anton’s Gun (cheap way of making an acronym AAA).

Atoms
Archetypes
Anton’s Guns

Now if you understand my ramblings and you’ve played or read about my plans about Link-Dead you will see that something does not fit here. A friend of mine when he played LD, always complained about the guns because he did not know anything about them (because they look like sci-fi weapons or some technology mash). He insisted I add standard guns (like an M14 or AK47). I thought he just needs to read the gun descriptions or fire it a couple times to get it. I gave up one day and introduced an AK for the Wasteminer faction. I had to admit this gun was much more fun to use than the others. I have no idea why because the parameters of the gun where exactly the same as the sci-fi one. But it just felt right. I knew what to expect from it. An AK is an AK. And that is how it works and you cannot bypass it. In order to have fun out of a model (which a game is) you need to know what to expect from it. Like a block in Minecraft, it’s cubic and you can put one on top of another. Nobody expects anything more and nothing less.

Movement is something I have been forever working on in Link-Dead. I just couldn’t get it quite right. There was always someone complaining that there is something wrong about it. But the hundreds of people that played it, noon could tell me exactly what it is. I made parkour movements with awesome animation. When I was making it I had great fun. I am really proud of this movement system. No other game has this and on specially built maps it is really fun to move. Yet when played on maps in multiplayer suddenly that enjoyment was lost. It wasn’t bad it just, with added shooting and other things, wasn’t as fun as just jumping around.

Only later after discovering these principles that I am writing here about I know what is wrong. The movement itself in Link-Dead is actually 2 or even 3 Atoms. And 2-3 Atoms could be a game in itself! I could actually strip everything down from Link-Dead and make a parkour only game without any shooting and it probably would have been more successful than it is now.

The next blow came after playing Ace of Spades. Someone posted in on the LD forums and I spent half night playing it. I haven’t done that for years. I was missing something. This game is fucking ugly and it is as simple as Manic Miner. What the hell? Now I knew for sure I have not simply grown out of enjoying games. The kid that enjoyed playing Doom for hours is still inside me. The same games are not here! I remember browsing through Ben’s blog (AoS creator) and people were bashing each other about Green Team’s superiority over Blue Team’s and drawing comics about the epic struggle between Green and Blue. It actually reminded me of what was going on with Soldat Red vs. Blue. And I remember thinking why the fuck am I trying to invent 2 super teams, with awesome looks and an epic storyline behind them when all you have to do is draw a blue guy and a green guy and leave the rest to the players imagination. People don’t need stories fed to them, they enjoy making them up by themselves!

I think at this point I realized I cannot continue working on Link-Dead. The fact is the games design is flawed. I have made an error. I assumed that the more stuff I add to the game the better it will be. The more detail the map will have the better atmosphere it will have. The more stuff you can do the better it will be. There are easily 10 or more Atoms/gameplay elements in Link-Dead that are there or I have planned. Heck, the hacker mini-game is a game in itself! It could easily be a hit casual game.

The problem is, so much stuff to do just confuses players. To the point they cannot state clearly what they think about a simple thing like the soldiers movement. I’ve made Prince of Persia + Soldat + Deus Ex + Thieve + Operation Flashpoint. Stacking games on top of each other just won’t work. All those games where successful because they relied on 2-3 atomic gameplay elements. PoP’s movement is a game in itself, whereas in LD it is “just” moving around. Soldat has:
1. Running/Jumping
2. Shooting
3. Flying
4. Grabbing stuff

This is very sad for me to say because I really tried hard, but if Link-Dead will ever appeal to a wider audience it should have at most those 4 elements. More simply won’t work. So not to make it Soldat 2 I would have to take some elements that are interesting in LD and base the whole game around it. For example:
1. Darkness + Flashlights
2. Traps
3. Killing
4. Simple movement

That’s all. No mech’s, no hacking, no doors, no scanners, no inventory. I have absolutely no faith now that it could work any other way. This is because I have tested my design principles.

When I realized AAA I just grabbed a piece of paper and started sketching the first game I could think of using the Principles. It turned out to be a social ant-type war simulator. It was a 2 team multiplayer game. It happens underground. Little people spawn out of eggs. You can pick 3 classes: a Digger, a Ladder guy and a Soldier which throws black balls. The purpose of the game is to cooperate with the other players to dig to the other teams base and steal their eggs. The eggs have to be brought back and given to the “King” so that he can eat them. There is no text in this game and I intended never to be any. The whole purpose of the game would be revealed by the King shouting at everyone, with a bubble icon representing that he want to eat eggs!

I called it Warmonger. I made a video of the game because I don’t have time to wrap it up and put for download:

I never played this game with anyone. The people you see in the game are bots. The AI is actually quite good. The diggers dig tunnels and when they can’t go up they request a ladder. Ladder people come and build ladders and so they go forward. They don’t dig blindly but find the fastest route across the map.

But what the hell am I talking about!? This isn’t the game that is right now on every gaming news website and which some guys tracked down me as being the author of the game.

I made Warmonger in 72 hours. I set a deadline like that. I wanted to work fast, just test out the idea. After that I was hungry for more. So again I sat with a pen and paper and this time I sketched another multiplayer game. This time I set the deadline for 2 weeks. I stopped posting on my blog. Didn’t tell anybody what I was doing, even my relatives and friends and just worked on this game. After 3-4 days I had the game more or less working and I knew I had something going on. I posted anonymously on a pixel artist forum and pitched my idea.

I pitched the game as a remake of this game:

An old SNES 2D RTS King Arthur’s World. One lucky pixel artist that goes by the handle Geti thought it’s a nice idea. He didn’t know what he was up for or who he was going to work with. I think I met the deadline of 2 weeks and the game went live on gamedev.net game developer forums. Anonymously. I did the same back in 2002 with Soldat. Now what happened from that point is a story of its own and I will write about it someday.

The reason I published the game by a different nick is that I wanted people to look at it with a fresh eye. Not associate it with Soldat or Link-Dead. I think a big problem with LD is that people can’t be honest with me. Nobody can tell me, look MM this sucks, do it again. That wouldn’t be honest also, because the game doesn’t suck. There are elements in it that are actually awesome so… I don’t know. I posted anonymously just to see what people say without any expectations.

During the last month King Arthur’s Gold got a huge crowd from Reddit and then every news site that has interest in indie games wrote about it.

Indie Games

PC Gamer

TIGSource

RockPaperShotgun

Kotaku

This never happened with LD. Only TIGSource is a stable fan and writes LD news. But this is not what depresses me most. The most depressing thing about LD is that nobody ever pirated it. And that is probably the single most telling fact about the game. Everything that is awesome gets pirated. And it’s not the security because there is none.

If you look at the KAG dev blog you will see that I am quite busy working on the game. And I will be in the nearest future, people really enjoy it and I want to give them more. I am not working on Link-Dead anymore. If I return to it, it will be a different game or I will work on it as a 2D technology show case. The map editor and the bump mapping and lighting are an impressive technology that nobody has. I want to make use of it.

Unfortunately we now live in an age where nobody cares about that. Pushing the limits of 2D graphics and tile mapped games doesn’t interest anybody other than a few die-hard fans and me. People are actually ecstatic about low-res pixel graphics (Minecraft). So until this fad dissipates I’m gonna join them. I see no point in fighting it.

Many of you of course payed for Link-Dead or to be correct donated. You payed whatever you wanted for the game and you got a working/playable game. I think whatever price you gave it is fair for a 3 map, 14 weapon + map editor multiplayer game. You can still play it with your friends or try to gather some people for a match. However I totally understand you might want to get your money back. If that is the case, please write to me (mm (at) thd.vg) that you want a refund and I will give it back to you. You might also want a second option which is: you will get the next game of mine in the future for free. It most likely will be a full version of King Arthur’s Gold (if it becomes payable in some form). It is yours, as a thank you for helping me survive and have food to eat for the last couple months.

Right now the lack of variation in content is more issue. Right now having full armour makes most weapons useless so you can choose two gameplay styles. Melee/Shotgun or Railgun. Anything else is useless against armoured players. Also everyone is carrying same items…

This quote by whitebear on the LD forums got me really thinking. It is true. I want Link-Dead to have tons of stuff to play with. Yet, it doesn’t matter that there is a lot of inventory to use. Everybody is using the same stuff. For example IED’s are a big part of Wasteminer weapons but nobody uses them because it is quicker to just kill with a pickaxe or shotgun. There is no reason for you in a competitive game to use something that is worse, so the best gun will always be used. The same goes for best load-out, best armor/class setup etc. Finally after a couple weeks of playing everybody will be using the same winning set. Even though there will be lots of content, variety will be killed. It’s like survival of the fittest. It doesn’t matter that a remote IED is cool, it is simply weaker (or harder to use) so it becomes extinct as a weapon of choice. For variety to exist, as well as for competitiveness and FUN to exist: the game must be balanced.

When designing the weapons I tried to give a lot of configurable variety for balancing a gun. So for example if a railgun kills with 1 shot, we can make it heavy (hard to aim) and load for a very long time. We can increase the loading time infinitely to balance the gun against the less powerful guns but by doing so we also decrease the fun we have out of a gun. A railgun can shoot once every 10 seconds but that is boring.

Balancing is a pain in the ass, as my experience with Soldat showed me. There are 14 weapons in Soldat and balancing them was years of effort (still it isn’t perfect). Right now there are 16 weapons in Link-Dead not counting all the options you can have with adding attachments. Add to that an armor system and device/character mutation system and we have an impossible to balance game. Wolfire Games wrote a blog about imbalanced games. That is cool for a singleplayer game, but a multiplayer game cannot ever be imbalanced.

So how do I get out of this situation? First idea that comes to my mind is “procedural balancing”. Let the computer do it for you.

For example: Statistical balancing could be done like this:

1. Gather online statistics about weapon kills from all servers
2. By some algorithm automatically nerf guns at the top of the list and pump up guns at the bottom of the list
This might work: Starcraft uses this system, the game that is most praised for its game balance. Counter-Strike Source also uses this to determine the price of guns in the shop. Gathering statistics however, does not *really* balance the game weapons themselves. It just balances whatever the players use most of the time. So if a weapon is just ugly and nobody uses it, the system will not see that and not balance it.

Another procedural balancing system that would fight that could be a balancing simulator:
Create test situations which could be repeated in an algorithm with bots such as:
a) shooting a guy from distance;
b) shooting from close;
c) shooting while moving
d) shooting an unarmored/armored guy etc.
And count how much kills a given weapon does.

These situations would be recreated thousands of times so that reloading time, jamming and everything that has statistical significance in a gun will be taken into account. The better the models for testing would be used the better the system would balance the guns. This solution seems reasonable but only if we don’t take into account individual playing styles. Many players will come up with their own ways of killing with a gun that would not be considered by the testing system (like shooting an M79 around a corner). To fight this we might join the first procedural balancing system with this one, or just tweak it by hand.

Even if this works, there is a second problem. What do we actually change when a gun is unbalanced? As I mentioned the railgun example we can increase loading times for a one-hit weapon. But can a procedural balancing system take into account how much fun a gun is? I don’t think so. Balancing the M79 in Soldat is the worst thing ever. For me and for many others it is the most fun gun to use in this game, anything done in order to balance it against other guns actually reduces the enjoyment of using it. So when making a game like this you end up with not only balancing the gun for a competitive environment, you are also balancing the game balance with game fun! As Arnold would say: Arghghhhhhhhhhh!

Now you see how hard my job is.

So any other solutions?
Yes, imbalanced weapons balanced through other artificial methods. These don’t change the fun of the gun or realism of a gun but limit how much you can use it. You might have thought of some of these of course: a point system, a real-money system, a research system, and a regulation system.

A point system is something like in Counter-Strike were you earn points and can buy guns from the shop with them. I don’t like this solution because:
1. It is boring
2. It rewards good players and penalizes the weak.
The better you are, the better guns you get hence the better you are? This is like a capitalist system in a game. The more money you have the more money you can make. This is I think the reason why Counter-Strike is a good game but isn’t very fun in the ROFL/LOL kind of way. People get too competitive and serious in this kind of environment. Not what I’m aiming for.

A real-money system is where you get only standard guns on start. If you want better guns you have to buy them with real money via micro-transactions. Many online FPS’s use this system. This works and is fairly profitable. If this was my only interest I would think about it. But all these online shooters suffer from the same problem which is: the more real-money somebody payed the better he is. Also none of these games is a fine work of art that will be remember, something which I want to do (if I’m desperate for money I will try it though, so be generous with the donations :D).

A research system is were you get some standard initial guns on start and besides the actual game itself there is a meta-game that takes care of the inventory you get. For example: your team can research new guns and weapons during the course of the game. I imagine this being fun, just like in an RTS game. For example one team researches a mine so now everyone can use this mine on the team. The other team to counter this will now steer research into a mine-detector. If one team invents a new pulse rifle, the other team can counter this with inventing a shield against this weapon. This sounds cool and if all weapons would be designed according the Rock/Paper/Scissor game theory it would work. Researching could be fun in itself by making the team hunt for resources to build the weapons or find technology and missing parts to construct them. You could also steal technology from your opponents. The thing to consider is: in matches lasting 15-30 minutes would this make any sense?

Pondering about this you might think why can’t I balance the current game just according to the Rock/Paper/Scissor theory? This works very well in RTS games were there are lots of units but in a game were you have one dude this doesn’t work. Lets imagine a simplified 1v1 game. Your opponent uses a standard rifle against you. You can apply the Rock here and put on armor to defend yourself against his lead bullets. You can kill him easily this way. After he respawns to counter your armor he selects a pulse rifle (that burns through armor). This time he kills you. So you take off the armor (because it is useless and heavy) and grab an energy shield that protects you from the rifle. This time you win. So he selects a standard rifle… and so on. You can see how silly this becomes. This is actually what is happening to a small degree in the current build of Link-Dead and why what whitebear wrote happens. In a game were there is more than 2 players it is impossible to play this inventory race game. Eventually everybody settles on one best load-out.

What’s left is a regulation system. If the point system was capitalism this is communism. It simply means that you regulate the amount of guns or clips that a team or player can pick. This is what Day of Defeat uses. In any given match there might be one sniper rifle per team, one Bazooka, two SMG’s and the rest are bolt-action rifles. This system works really well and is fairly realistic (in real wars teams are structured like this). It also provides variety for different classes to exist (not everyone picking bazookas and turning WW2 into Quake). The thing I don’t like about this system is that if you want to be a sniper you can’t be a sniper! You either have to wait, try to be lucky or try to be fastest to pick the class after round restart. It is a little weird to compete with your teammates for resources. Not to mention the times when a bozo picks a sniper and doesn’t actually do anything with it and everybody is just screaming at him to give up the sniper role!

A variety of the regulation system is to just limit ammo clips. So if you have a strong railgun, there are only 10 clips you can use in a given match. This is also very realistic since in real wars you are limited to what the headquarters give you. I can’t think of any downside except the fact that probably the first half of a match everybody will be shooting each other with explosive nukes and the second half they will fight with sticks and stones (like in that Einstein quote about WW3). But this isn’t such a downside because it gives opportunity for clever tactics. Such as waiting for everybody to use all their good guns and then bring out the Ol’Painless and mow all the poor bastards with hand guns.

Writing this cleared up some of my thoughts. I think I am biased now toward a regulation system with limited clips. Might add a spice of research to it for example. Nothing for sure though, I need a discussion. Anything I missed? What are your thoughts? Very much appreciated. Remember we are making the best game ever here, we want lots of cool stuff to use but we want it be competitive and balanced too.
Forum link is here.

If you have a good idea that doesn’t involve hit chances please post, I am very interested.

The last picture shows why this approach is correct from a realism point of view. It is still to be considered if it is fun and has gameplay value. I’ll definitely try it out in Link-Dead.

I’m imagining how Link-Dead would be as a single-player game. The way I would do it would be totally different than anything that ever was seen on the gaming market. Half-Life 2 has episodic content. I would do user-created experience-based content.

The primary motivation for this is my understanding of what we want from a computer game. Especially as I got older and I see other older players (more than 20 yrs) and what they want. For example my brother has a PS3 and the only game he plays is Fight Night. That is because he just wants to pop-in for 10 minutes and punch a dude in the face. That’s all there is to it. We don’t have time for more. I am personally frustrated by games that take long to load, make me watch cinematics, make me become involved in a story, make me start slow and do progress in a game. I DON’T HAVE TIME FOR THAT. I just want to get in a game and shoot some dudes in the face for 15-30 minutes a day. That’s all. Soldat partly solves this problem, because it is simple and allows for exactly that. But I have a better idea that would allow the player to become more or less involved in a story making the game not so repetitive.

Lets say this is a game like Link-Dead but not multiplayer (so future war, destroyed cities, machines etc.). The basic idea from which everything would spawn is: you are a regular soldier in a war. You log in a map and immediately become part of the story happening there. Many different stories like: you are lost alone and have to find your way back to base; you are leader of a squadron; you are a scout; you are commander of a mech robot; you are part of a squad in the role of a hacker etc. These stories/maps would be user-created allowing for infinite possibilities. These would be 5-30 minutes experiences which you just take and leave.

When I’m playing any game my motivations are usually as: today I feel like shooting dudes in the legs; today I feel like being a large robot and destroying everything; today I want to stay quiet and snipe. Some games provide these possibilites, but most fail, or the experience is frustrating to get to. Like you have to conquer 10 maps to get to the point were you want to. For example Half-Life. The most interesting part was fighting the grunts and black ninjas (or whatever they were). I played those maps repeatedly many times because they were so cool. And the AI allowed for many possibilites of resolving the situation. This is what I would want. Little experiences driven – on the micro-scale by very good AI and on the macro-scale by well planned human created story.

A story is just a map, your place in it (your role) and different enemies and situations to solve. The stories would be created by something like a Game Master. An idea from RPG games (not cRPG). If I would be brave I would do it like this: a Game Master would not just create a map. He would be like a game server host and admin of the server plus more. He would be looking over his game in real-time. That means watching as human players join and play and modifying the game in real-time. A good Game Master would add or subtract gameplay elements as he feels the player needs. For example if he sees a player wandering the map, becoming obviously bored – place an enemy somewhere nearby or place a trap for him to fall into and wake him up. If the game is too hard for someone – place a medkit or a cool weapon that he can find. Shoot an artillery strike just for spicing things up or make the players weapon jam if he is overusing it.

I have also more ideas thanks to the TwitPic bugtracker I made in the Technology Test. Instant feedback would be a core functionality in this game. Things that are bad would be removed quickly and things that are good would be emphasized. Good maps and Game Masters would be promoted and weak ones would be weeded out.

Oh yes and if I would be super-brave this would be a co-op game. Meaning that if you are playing some other players can join. Either as your teammates, separate from you or enemies. Adding randomness to the whole experience (you would never know if the enemy soldier you are fighting against right now is AI driven or a human).

That was my flash of inspiration. This is my dream game.

A lot of times while playing Day of Defeat I wasn’t really playing for the purpose of killing as much enemies as I can, I played to survive for the longest time possible. I think my personal record was nearly an hour, being a sniper with more than 20 kills. Yes it involved a lot of hiding and camping but it was worth it. Trying to stay alive is as much fun as killing.

Now how do you bring this into ordinary gameplay, in a game like Soldat for example?

You must somehow force the player to be careful by restricting the time he has for gameplay. This works fairly well in Counter-Strike or Soldat survival mode. But what if we take this one step further?

My first idea was a once per lifetime event. You download the game. You enter the battle. If you die the game quits, erases itself and you can’t ever play the game again in your life. It would be as if you actually were born and died in the game.

For me it sounds brilliant… but maybe not for the rest of the gamers.

So the next idea is… Weekly Battles.

Once per week at a given time there would be a battle. This would be the actual game happening. 10000’s of people would join the server and fight until they die. Some men would die in the first few minutes, others would fight for a couple days. This would be possible because of the fear of death. There would be no Rambo spraying and running into the enemy base. Believe me if you had a chance to play the game once per week you would value your life enormously. Half of the army would probably stay in base just for safety. There would probably have to be a special sergeant class that kills deserters and sends them to battle!

If this idea seem bizarre to you, well think about TV shows. I think you don’t have a problem with waiting a whole week for your favourite sit-com? And they last mostly 30 minutes! This would probably be also the average time of living in this game. But imagine the intensity! You would remember every minute of it for the rest of the week! A gaming serial!

A single battle would carry on for hours. People would not sleep or hide somewhere and go to sleep. When they would wake up the battle would still be there. Each player carefully planning every step and discussing it with his team mates. The players that would survive would become veterans. They would wear special badges and everybody would fear them. If you would see a veteran you would flee for sure! But imagine the glory if you killed one of them!

Of course in-between battles, during the week, you would get to train. There would be special boot camps which you would have to sign up for. You would train on special maps and shoot with paint balls. Sort of like the noob missions in America’s Army. But that would just be so that you can prepare yourself during the week for the actual fun! But if you suck too much during the boot camp you would be punished. Like if you did not show up for training you would get degraded and then when the actual battle would happen you would be sent to front lines as cannon fodder. Or if you would be really bad you would be banned from the battle and wait for next week to pass the boot camp. The drill sergeants in the boot-camp would of course be veterans. See how this all works elegantly?

This idea nicely blends with another fantasy of mine. When you play any multiplayer game you don’t really know who you are playing with in your team. You’re supposed to know each other, because the game would be better, teamplay would be more smooth, but there is no real possibility for that, the game is too quick. This is partly solved by clan wars, but again it doesn’t matter if your teammate will die or not. If he does he will respawn in 5 seconds and that’s it. There are no feelings here.

But gaming is all about feelings.
So what if you could bring this feeling into the game…

You land in a marine boot camp with 10 other soldiers. You train with them all week, the same guys everyday. During that week you get to know everybody. You bond with each other because you have a drill sergeant which you all hate. You become buddies. Let’s imagine during this week you acquire a friend named Chuck. You fight with each other really well, you have great rapport and do all the team-based missions together. You’re both becoming stronger everyday. Finally the day of the battle arrives. You, Chuck and all your team mates disembark on enemy territory. After days and days of training you stand on the ground looking for enemies… then within seconds out of nowhere… a giant robot stomps right next to you crushing Chuck into little pieces! NOOOOOOOOOOOOOOO!!!!

You fight furiously, half of your team is gone. After minutes the fight is over. There are no survivors except you. All the people you knew are dead and you won’t see them ever again (well for a week at least). You look around and you notice that Chuck isn’t really dead. He’s got 1% health and he’s lying in a pool of his own guts. You quickly rush to him with a medikit. You try to revive him. Shit if he dies the game ends, all this week will be a waste! While you’re patching him he mumbles to you “Tell my wife that I…”. You reply “No, hang on, you will tell her that!”. “Tell my wife…. ehhhhh…*gargle*gargle*”. “Hang on Chuck! Chuck!? NOOOO!!!! YOU DON’T EVEN HAVE A WIFE!”. And then another giant robots comes from around the corner and stomps on both of you. End of game.

Do you see this? This level of intensity can’t ever be achieved with normal 5 second spawn times or rounds. This intensity comes from valuing life!

If you have other ideas for this let me hear them. The idea is to make the player value his life and the life of his teammates for intensity, fun and adrenaline! I want to see this one day.

The reason that FPS games are nearly impossible to mod now is the amount of work needed to complete a mod. Mod teams consist now of several or even dozens of artists. Maps and models are created in days. More polygons, higher resolution textures, more demand for realism. If you have money you can do this, but if you’re making games for fun, it is not fun anymore.

So where will all the talented and motivated people go now?

First let’s see why did the mod community focus on first person shooters?
It all started with Doom. Doom is the father of 3D, multiplayer and modding as we know it today. It started with a simple file format called .WAD. What it was, was simply a zip archive with all the game media in it. Graphics and sounds were trivial to mod. The map format, I believe was hacked and a dozen of map editors spawned. Later on we could even find EXE editors which hacked Doom network code and added jumping and new mouse look features. The ease of modifying things and quickly seeing the results was very attractive.

Then arrived Quake with an embedded scripting language called Quake C. This allowed for almost infinite creativity. These were the times were anyone could come in and make a decent mod. Even by himself. Models had 100-500 polygons, the animations didn’t matter. Textures had a size of 64×64 instead of 1024×1024. Maps were made out of simple sectors instead of being modelled and lightmapped for a week in expensive 3D software. It was a virgin territory, ideas flew in just like that. It was the golden era of modding which finally ended with Counter-Strike and Day of Defeat built on the Half-Life engine which was an improved version of Quake 2. Both mods went commercial and are making money now for Valve. Then it stopped, no big mod that everyone would know of after that.

Now I am here to bring those times back. All my efforts with the new game besides making it the best game ever made go into modding capabilities. I seek new technologies to accomplish this mighty task. I won’t say yet what it is but I will write briefly about how maps will look in my new game:

To state it as simply as possible. The maps will be 100% painted.

No more polygons (no more polybugs:p).
No more sceneries except for dynamic elements.
Collisions will be made out of pixels similarly to Liero or Worms, but much much better using a totally different approach.

What will be the map editor? Your favourite paint program (Photoshop, Paintshop Pro?).

The artist will simply paint the map, specify the layers, save them and then convert the map, add spawn points, dynamic elements etc. in the in-game editor (the beginning of the implementation can be seen in Crimson Glory).

This is a never seen before technology in a 2D game. It is not tilemaps and it is not polygons. As I said it looks similar to Worms or Cortex Command but all games using that technology are limited to the size of the maps because bitmaps get really big and stop fitting in system memory. The size will not be a big deal here. Also those games suffer from collision inaccuracy, here it will be perfect.

So goodbye PMS format. No more polycount and scenery counts.

Of course I will write more about this in the future, telling all the details, I’ll probably post some fairwell pictures to the PMS maps working in the new engine cause I have that working with a walking character. So stay tuned.

This spawned after my brother proposed it, we discussed it intensely and I realized I had a very similar idea years ago while playing Day of defeat, but I never thought of doing that in Soldat or in any other game. Now it seems very intriguing.

The idea is called seamless servers, as in seamless textures (textures that when joined in groups have no apparent beginning or end). Let’s discuss the simplest case based on this picture:  

Three servers (A, B, C) are joined together creating a seamless server. For the outside world they are seen as one. For example if a player opens the server browser this junction is seen as one server, let’s call this a game node. Players instead of joining a single server join a game node.

Server A runs ‘map A’ (green rectangle area). Server B runs ‘map B’ (blue). Server C runs ‘map C’ (red). These maps are constant and don’t change unless all three are changed at once to a different set of maps ABC.

On joining, as a red team member, the player arrives on server A. Through game progress he can move between servers, as in the example he can move forward according to the game flow. He starts on the left side and when reaching the right edge of the map he can switch servers and move to server B (blue rectangle). This is done instantly, because all servers know about the player, so there is no long connecting, and the map size is really small so they can be held in memory.  The player reaches the edge, zap, he is on the other server (map).

What to do with this? I have several cool gameplay ideas and features for this:

• Conquest mode

This was the idea I had with Day of defeat. Imagine a large city. A city to be conquered by an army must move from sector to sector and win each of them. A city is a bunch of sectors, each sector is a map, each map is running on a separate server, all joined together in a seamless server. On the image above you can see red and blue dots, these are flags to be conquered. If all flags are taken on a map the players spawn and play on the server where the conflict is not yet defined (server B). The game is ended if a team conquers all sectors.

• RPG mode

Role-playing and long hours of play on one server will make a lot of sense with this. This will be one big world with different maps. Players will move between servers according to their experience. So at start a player would join a noob map and move from there. This would be exactly as a Massive-Multiplayer game but of course much cooler.

• Goodbye lobby server

Imagine a multiplayer game without a server browser. Anything like that out there? I don’t think so. Let’s say every single server in the world would be joined creating one game node, or several, for example one for each continent ‘(to solve ping issues; one game node for Europe, one for North America etc.). In each node there would be hundreds of servers. The way you would travel between servers would be through portals. Each server would link to other servers and in different parts of maps there would be little portals through which you could instantly get to another server. Of course these portals would have little info drawn on them like ping, number of players, map would be seen just by looking through the portal. Imagine chasing each other not just on one map but through several servers and maps. You could hide on a different server or go help your buddy on a different one, all without exiting the game once, all realtime.

Of course there is much much more to this. I won’t write about the technical aspect of this because it is boring. The thing that matters is that this can be done. Definitely I want to try this out in my next game. In the meantime it can be done in Soldat.

What a seamless server needs is:
– a powerful host to setup 2 or more servers.
– specifically designed maps which would act as one (or just split an existing one)
– a script which would detect if a player reaches the edge of a map
– disconnect the player and make him join again on the next server (this can’t be done but I can make it in 5 minutes for version 1.4.2 (feature added))

What do you think?